Deep Learning Crash Course: Introduction

April 29, 2018

Transcript of the video

There are many tasks that are easy for humans to perform but hard to articulate how we do it. For example, how do we recognize faces or understand the verbal language? We do it effortlessly, but it's hard to transfer that knowledge to machines as a set of instructions.

What if we could teach computers by example? Instead of providing them with a comprehensive set of rules, we could show them some examples so that they can understand how the world works. Machine learning addresses these problems.

Deep learning is a term used for a subset of machine learning models that use several layers of learnable parameters. This allows for building a hierarchy of simple features that can express complex concepts.

The first layers learn simple features, and the representation becomes gradually more abstract as we move closer to the output layer.

One thing that makes deep learning interesting is that we don't tell the model what features to extract at every layer. All we need to do is to provide input and output pairs. Then, the model figures out what kind of features would be useful to learn a mapping between these two. It's not always that simple, but it's still fascinating that deep models can learn hierarchical representations from data.

Why is deep learning great? Because it works well. Is it the ultimate answer to all problems in AI? No, but it seems that it will be around for a while until the next paradigm shift in this field.

So, it's useful to know how deep learning works and how to get most out of it. That's what we are going to be doing in this series. We will learn the fundamentals of machine learning, with a focus on deep learning. We will talk about where to find data, how to build models that can process data, and generate data as well.

So, stay tuned and don't forget to subscribe for more videos. The videos will be short. You can finish the entire series in one afternoon if you wish. You can find them right here in this playlist. Alright, welcome on board, I will see you next time.